Call me crazy...new cooling method

prankstar008 · Dec 2, 2004

Has anybody ever thought of soldering wires to the individual pins on a processor and then soldering the other ends of thie wires directaly to the mobo. This would allow the processor to operate several feet (more or less) away from the mobo thus making submersive cooling easily obtainable. (IE a small box with a non condictuve chilled fluid running through it, with the keyword being small)

Is it even possible for the processor to operate like this..?.or would the resistance of the wires be a problem? Has anyone ever thought of this before? Am I nuts? Forget that it would make the mobo and processor combo permanent and void all sorts of warentees.

ps2cho · Dec 2, 2004

Im not pro at that...but it sounds llike an extremely good idea !

theMonster · Dec 2, 2004

Prakstar008 said:
Call me crazy

You're crazy.

p.s. Not really, but I shudder to think of the nightmare that this would be for what little benefit it would provide.

gastro54 · Dec 2, 2004

I bet the added resistance, inductance and whatever else would seriously screw it up

Aphex_Tom_9 · Dec 2, 2004

i highly doubt a 'remote socket' idea like that could work. interference/induction, signal strength...the farther it goes the more possibility of screwing up a signal in transit exists. that's why you dont see 5 foot long IDE cables...

AEsnowboarding · Dec 2, 2004

Hey if you wanna solder all 939 pins on a A64 go for it :thup:

prankstar008 · Dec 2, 2004

all those things are sorta what i thought....id just been wondering about this for a while

prankstar008 · Dec 2, 2004

there arent really 939 pins on a 939 socket chip are there?....i just build a system w/ a socket 754 (wtvr athlon 64's are for poorer people) and it didnt look like 745 pins?

Voodoo Rufus · Dec 2, 2004

The main problem with this is that the timing between data pins is EXTREMELY tight, hence why you see zig zag traces on the boards for the memory, to help data arrive at the same time to the chipset. So the short answer is no, it wouldn't work except for really low clock rates.

Aphex_Tom_9 · Dec 2, 2004

prankstar008 said:
there arent really 939 pins on a 939 socket chip are there?....i just build a system w/ a socket 754 (wtvr athlon 64's are for poorer people) and it didnt look like 745 pins?

oh yes there are...

just look at that...each row of pins has about 30, there are 30 rows, 30*30=900, which is only 39 pins away from 939...so without actually counting them, it's a pretty damn safe bet it's 939

iNterBeiNg · Dec 2, 2004

Just think each wire would have to be exactly the same length too..

JaY_III · Dec 2, 2004

and what would be the point of having the CPU run far away from the Motherboard?
If you plan to submerage you system, thier is not reason why you wouldnt want to do that for all non moving parts.
3M has been making a product for that for many years now

Aphex_Tom_9 · Dec 2, 2004

jAY said:
3M has been making a product for that for many years now

Flouroinert :drool:

mad expensive tho

JaY_III · Dec 2, 2004

Aphex_Tom_9 said:
Flouroinert
mad expensive tho

Its only like what, 350-500USD a gallon.

Voodoo Rufus · Dec 2, 2004

Price probably depends what temp range you want, but cool stuff.

man_utd · Dec 2, 2004

That, and this idea has been discussed many times, and never done successfully. Just not worth it.

prankstar008 · Dec 2, 2004

and what would be the point of having the CPU run far away from the Motherboard?
If you plan to submerage you system, thier is not reason why you wouldnt want to do that for all non moving parts.
3M has been making a product for that for many years now

Its only like what, 350-500USD a gallon.

That would be the point of this whole idea...you would only need a small enclosure requiring only a small ammount of liquid....but i guess its damn near impossible

JaY_III · Dec 2, 2004

prankstar008 said:
That would be the point of this whole idea...you would only need a small enclosure requiring only a small ammount of liquid....but i guess its damn near impossible

then why dont you just ude phase cooling on just teh CPU?
Die-electric grease is cheap, and that takes care on condensation.

The Design aspect of what you are thinking is a nightmare, it makes no sense.
Now if you did manage it somehow,with the price it would cost the Flouroinert would still be a cheaper way to do things

prankstar008 · Dec 3, 2004

then why dont you just ude phase cooling on just teh CPU?
Die-electric grease is cheap, and that takes care on condensation.

The Design aspect of what you are thinking is a nightmare, it makes no sense.
Now if you did manage it somehow,with the price it would cost the Flouroinert would still be a cheaper way to do things

I totaly agree with everythign everyone has said...except "thoreticaly" this would be cheaper in the long run because running a compressor uses a lot of energy..but you are right in the fact that phase change cooling is more logical and acctualy obtainable.

[EDIT] I just re-read what I wrote....and i basicaly just said what you just said (huh?)....never mind[/EDIT]

Aphex_Tom_9 · Dec 3, 2004

lol, it's okay, i thought of it once..."what if you could spread a computer across a whole room, like different parts in different places", but then realized you cant keep signal intergrity that way.

Call me crazy...new cooling method

Member

Member

Disabled

Member

Member

Member

Member

Member

Powder Junkie Moderator

Member

Member

Senior of BX

Member

Senior of BX

Powder Junkie Moderator

Member

Member

Senior of BX

Member

Member

Similar threads